Register to reply

Hermite Polynomials

by Domnu
Tags: hermite, polynomials
Share this thread:
Domnu
#1
Aug15-08, 08:43 AM
P: 178
Since the Hermite Polynomials are orthogonal, could one state that they span all polynomials [tex]f[/tex] where [tex]f : R \rightarrow R[/tex]? This would be EXTREMELY useful for the harmonic oscillator potential in quantum mechanics...
Phys.Org News Partner Science news on Phys.org
'Office life' of bacteria may be their weak spot
Lunar explorers will walk at higher speeds than thought
Philips introduces BlueTouch, PulseRelief control for pain relief
Jimmy Snyder
#2
Aug15-08, 01:14 PM
P: 2,179
Quote Quote by Domnu View Post
Since the Hermite Polynomials are orthogonal, could one state that they span all polynomials [tex]f[/tex] where [tex]f : R \rightarrow R[/tex]? This would be EXTREMELY useful for the harmonic oscillator potential in quantum mechanics...
No. If you think of the two vectors:
(1, 0, 0) and (0, 1, 0)
They are orthogonal to each other, but they don't span.
on the other hand, consider the three vectors:
(1, 0, 0), (1, 1, 0), (1, 1, 1)
They span, but they are not orthogonal to each other.
Therefore, knowing that the Hermite Polynomials are orthogonal is not enough to show that they span. You would need to prove that.
Avodyne
#3
Aug15-08, 01:28 PM
Sci Advisor
P: 1,232
It doesn't follow from orthogonality alone, but it is nevertheless true.

Domnu
#4
Aug15-08, 02:00 PM
P: 178
Hermite Polynomials

Quote Quote by jimmysnyder View Post
No. If you think of the two vectors:
(1, 0, 0) and (0, 1, 0)
They are orthogonal to each other, but they don't span.
on the other hand, consider the three vectors:
(1, 0, 0), (1, 1, 0), (1, 1, 1)
They span, but they are not orthogonal to each other.
Therefore, knowing that the Hermite Polynomials are orthogonal is not enough to show that they span. You would need to prove that.
Fair enough... okay so I guess a better argument would be that the Hermite polynomials are infinite in number and are all orthogonal, so they span all functions?
Useful nucleus
#5
Aug15-08, 02:07 PM
Useful nucleus's Avatar
P: 264
They are useful to express any function defined on the range on which they are orthogonal.
Other orthogonal polynomials will be useful to express functions on the range on which they are orthogonal (e.g. Legender polynomials on [-1,1].
morphism
#6
Aug15-08, 02:21 PM
Sci Advisor
HW Helper
P: 2,020
Quote Quote by Domnu View Post
Fair enough... okay so I guess a better argument would be that the Hermite polynomials are infinite in number and are all orthogonal, so they span all functions?
No, that doesn't make any sense.

In what sense are you using the word "span"?
Domnu
#7
Aug15-08, 02:27 PM
P: 178
I'm using the word span in the sense that any polynomial of the form

[tex]f(\xi) = a_n \xi^n + a_{n-1} \xi^{n-1} + a_{n-2} \xi^{n-2} + \cdots + a_1 \xi + a_0[/tex]

can be written as a linear combination of [tex]H_0, H_1, \cdots, H_n[/tex]. I see I had a typo in my last post... I meant that the Hermite polynomials could span all polynomials (real).
Jimmy Snyder
#8
Aug15-08, 05:24 PM
P: 2,179
Quote Quote by Domnu View Post
Fair enough... okay so I guess a better argument would be that the Hermite polynomials are infinite in number and are all orthogonal, so they span all functions?
No. For instance, the following are mutually orthogonal and infinite in number, but they don't span,
(0,1,0,0,0,...)
(0,0,1,0,0,...)
(0,0,0,1,0,...)
...
Dick
#9
Aug15-08, 05:35 PM
Sci Advisor
HW Helper
Thanks
P: 25,235
Quote Quote by jimmysnyder View Post
No. For instance, the following are mutually orthogonal and infinite in number, but they don't span,
(0,1,0,0,0,...)
(0,0,1,0,0,...)
(0,0,0,1,0,...)
...
True. But it's pretty clear the Hermite polynomials span. H_n(x) contains a term containing x^n and none of the preceding H's do. You can formally prove it by induction, but it's actually pretty obvious if you imagine how you would go about expressing a given polynomial in terms of Hermite polynomials.
Jimmy Snyder
#10
Aug15-08, 06:00 PM
P: 2,179
Quote Quote by Dick View Post
True. But it's pretty clear the Hermite polynomials span. H_n(x) contains a term containing x^n and none of the preceding H's do. You can formally prove it by induction, but it's actually pretty obvious if you imagine how you would go about expressing a given polynomial in terms of Hermite polynomials.
My guess is that Domnu would benefit by doing the induction.


Register to reply

Related Discussions
Hermite Polynomials. Calculus 4
Hermite Polynomials Advanced Physics Homework 0
Hermite Polynomials Advanced Physics Homework 0
Hermite Polynomials Calculus & Beyond Homework 2